cell-AAP
Utilities for the semi-automated generation of instance segmentation annotations to be used for neural network training. Utilities are built ontop of UMAP, HDBSCAN and a finetuned encoder version of FAIR's Segment Anything Model developed by Computational Cell Analytics for the project micro-sam. In addition to providing utilies for annotation building, we train networks using FAIR's detectron2 to
- Demonstrate the efficacy of our utilities.
- Be used for microscopy annotation of supported cell lines
Cell-line specific models currently include:
- HeLa
- U2OS
Models have demonstrated performance efficacy on:
- HT1080 (HeLa model)
- RPE1 (U2OS model)
We've also developed a napari application for the usage of these pre-trained networks.
We highly recommend installing cell-AAP in a clean conda environment. To do so you must have miniconda or anaconda installed.
If a conda distribution has been installed:
-
Create and activate a clean environment
conda create -n cell-aap-env python=3.11.0 conda activate cell-app-env
-
Within this environment install pip
conda install pip
-
Then install cell-AAP from PyPi
pip install cell-AAP --upgrade
-
Finally detectron2 must be built from source, atop cell-AAP
#For MacOS CC=clang CXX=clang++ ARCHFLAGS="-arch arm64" python -m pip install 'git+https://github.com/facebookresearch/detectron2.git' #For other operating systems python -m pip install 'git+https://github.com/facebookresearch/detectron2.git'
- To open napari simply type "napari" into the command line, ensure that you are working the correct environment
- To instantiate the plugin navigate to the "Plugins" menu and select "cell-AAP"
- You should now see the Plugin, where you can select an image, display it, and run inference on it.
If running inference on large volumes of data, i.e. timeseries data >= 300 MB in size, we recommend to proceed in the following manner.
- Assemble a small, < 100 MB, substack of your data using python or a program like ImageJ
- Use this substack to find the optimal parameters for your data, (Number of Cells, Network confidence threshold)
- Run Inference over the volume using the discovered optimal parameters
Once inference is complete the following colors indicate class prediction
- Red: Non-mitotic
- Blue: Mitotic
For analysis purposes, masks in the semantic and instance segmentations have the following value mapping: Semantic
- 1: Non-mitotic
- 100: Mitotic
Instance
- $2x$: Non-mitotic
- $2x-1$: Mitotic
Version:
- 0.0.9
Last updated:
- 20 February 2025
First released:
- 28 May 2024
License:
- Information not submitted
Supported data:
- Information not submitted
Plugin type:
GitHub activity:
- Stars: 0
- Forks: 0
- Issues + PRs: 0
GitHub activity:
- Stars: 0
- Forks: 0
- Issues + PRs: 0
Python versions supported:
Operating system:
- Information not submitted
Requirements:
- napari[all]>=0.4.19
- numpy==1.26.4
- opencv-python>=4.9.0
- tifffile>=2024.2.12
- torch>=2.3.1
- torchvision>=0.18.1
- scikit-image>=0.22.0
- qtpy>=2.4.1
- pillow>=10.3.0
- scipy>=1.3.0
- timm>=1.0.7
- pandas>=2.2.2
- superqt>=0.6.3
- btrack>=0.6.5
- seaborn>=0.13.2
- openpyxl>=3.1.4
- joblib>=1.0
- scikit-learn>=0.22
- cython<3,>=0.27